Taiko Audio SGM Extreme : the Crème de la Crème

It's been a while since I compared dCS' Mosaic to the Extreme/Roon or Extreme/XDMS. Easy enough to do and I'll give it decent trial streaming from Qobuz this evening when I sit down to listen to music after dinner. I'll let you know what I hear.

Later,

Steve Z
Much appreciated
 
  • Like
Reactions: oldmustang
not to derail the thread, but i recently tried an antipodes oladra and found their connection over toslink with the word clock to be extremely enjoyable and in many songs better than their usb to the proisl connection. I'm not sure if this is from the server being designed to work better over toslnk/aes/is2 vs usb or not. Thus my question about which input is best suited for MSB with taiko. i also use the DD with my system fwiw
All Antipodes (Oladra and any other recent model) outputs except USB are derived from a separate proprietary "Reclocking Board" that not only reclocks but is also in large part responsible for the server's signature free-flowing and organic nature. TOSlink in itself is strictly speaking an inferior format that tends to round transients. But versus the relatively cleaner/more technical-sounding USB implementation in Antipodes servers, the net effect can indeed still be more pleasing and thus subjectively better.
 
  • Like
Reactions: John T and MarcelNL
Toslink is based on SPDIF, I have yet to hear a good implementation.
 
Toslink is based on SPDIF, I have yet to hear a good implementation.
That is correct and I would certainly endorse using Coaxial S/PDIF over TOSlink. AES/EBU, by the way, is also very similar to S/PDIF, but just like S/PDIF, it can work very well, depending on the implementation. For Taiko Audio servers, the opposite is true and USB works best. USB is theoretically ideal, but it takes a lot of effort to actually make it ideal. It's all in the implementation.
 
Last edited:
  • Like
Reactions: oldmustang
Toslink is based on SPDIF, I have yet to hear a good implementation.

Indeed Toslink should be able to perform competitively to SPDIF but its implementation is often done in the cheapest possible way.

A further consideration is sample rate support which per interface is limited to:

Toslink : 96KHz, no native DSD
SPDIF : 192KHz, no native DSD
AES/EBU : 192KHz, no native DSD
USB : currently supports but technically not limited to : 1536KHz, native DSD up to DSD2048

Then there is Ethernet ofcourse but that’s a different type of interface as it requires a “streamer / endpoint” to reside inside your DAC, to quote @MarcelNL , “I have yet to hear a good implementation”. There are more advanced interfaces running on Ethernet, like Ravenna, but these are specialised interfaces and exclusive to just a handful of DACs, these can also be used for high bandwidth multichannel systems.

The new kid on the block will be our own interface launching in the next few months, designed from scratch, both hardware, firmware and drivers, supporting any sample rate. Which is CPU direct over PCIe to I2S (“DAC language”). It can be clocked from both source as destination.

Why we went through the elaborate process of designing this is all the “off the shelf” solutions, which are the “engineering blocks” everybody uses, are aging and development on those have stopped. Further to this all of them have limitations in one way or another, they’re standard building blocks needing specific surrounding circuitry to operate which you have no choice in but to follow. Things like jitter reduction, reclocking, very expensive clocks etc are really just bandaids to reduce the negative impact of the existing issues in these standard circuits/protocols.
 
Last edited:
I hope the format will become widely available, the PF I2S bridge is getting a bit long in the teeth (although it is one of the best generic sounding systems IMO)
 
I hope the format will become widely available, the PF I2S bridge is getting a bit long in the teeth (although it is one of the best generic sounding systems IMO)

Yes, and also a good example by the limitations imposed by having to use off the shelf solutions on which there’s no further development. Like a 192KHz sample rate limitation and ancient drivers.

In sheer size High-end audio has become a smaller market and R&D costs have increased dramatically. Hence large companies no longer invest in developing products in the “niche market” of high-end audio as the ROI is not there. We now have a very limited choice in DAC chips supporting a wide range of samplerates for example, really only 3, ESS Sabre, Rohm and fortunately AKM is back in business. Next to that availability of the high-end versions of these chips is very low, you have to order 6-12 months in advance (Rohm & AKM)..

The high-end audio companies which do continue to design improved D/A solutions keep these proprietary to themselves and sell their products at a premium to be able to get a ROI on that.
 
Indeed Toslink should be able to perform competitively to SPDIF but its implementation is often done in the cheapest possible way.

A further consideration is sample rate support which per interface is limited to:

Toslink : 96KHz, no native DSD
SPDIF : 192KHz, no native DSD
AES/EBU : 192KHz, no native DSD
USB : currently supports but technically not limited to : 1536KHz, native DSD up to DSD2048

Then there is Ethernet ofcourse but that’s a different type of interface as it requires a “streamer / endpoint” to reside inside your DAC, to quote @MarcelNL , “I have yet to hear a good implementation”. There are more advanced interfaces running on Ethernet, like Ravenna, but these are specialised interfaces and exclusive to just a handful of DACs, these can also be used for high bandwidth multichannel systems.

The new kid on the block will be our own interface launching in the next few months, designed from scratch, both hardware, firmware and drivers, supporting any sample rate. Which is CPU direct over PCIe to I2S (“DAC language”). It can be clocked from both source as destination.

Why we went through the elaborate process of designing this is all the “off the shelf” solutions, which are the “engineering blocks” everybody uses, are aging and development on those have stopped. Further to this all of them have limitations in one way or another, they’re standard building blocks needing specific surrounding circuitry to operate which you have no choice in but to follow. Things like jitter reduction, reclocking, very expensive clocks etc are really just bandaids to reduce the negative impact of the existing issues in these standard circuits/protocols.
Will your new dac render upsampling
devices redundant or not necessary?
 
Will your new dac render upsampling
devices redundant or not necessary?
The intended product will be modular and offer either Digital or Analog outputs. With the former, you can specify "standard" digital interface formats that still allow the use of any DAC and upsampler you like. You can also specify the new Taiko interface format which will limit the connectivity to manufacturers that have also implemented this new format. The analog-output DAC version, finally, is a closed system.

The Modular Daughter Board will provide one specific function from an array of options such as:
1. A DAC with Analog output
2. Generic S/PDIF, AES/EBU, I2s, or other digital formats
3. Custom output like MSB Pro ISL or Dual AES/EBU for DCS
4. New Taiko proprietary interface
 

Attachments

  • TACDA & TACDD DAC & Interface Summary_v22.pdf
    428.7 KB · Views: 23
Will your new dac render upsampling
devices redundant or not necessary?

Upsampling has more to do with DAC filtering, that difference remains. The advances are more in the interface, where uncompressed PCM 24/768 will utilise about 0.00000022% of the interface bandwidth. Therefor higher resolutions are less impacted by interface “limitations” then before. IOW the difference between 16/44.1 and 24/768 becomes more or less negligible from an interface perspective.
 
Upsampling has more to do with DAC filtering, that difference remains. The advances are more in the interface, where uncompressed PCM 24/768 will utilise about 0.00000022% of the interface bandwidth. Therefor higher resolutions are less impacted by interface “limitations” then before. IOW the difference between 16/44.1 and 24/768 becomes more or less negligible from an interface perspective.
Will there be an option to connect to a Chord scaler which is usb?
 
Will there be an option to connect to a Chord scaler which is usb?

No then we’d be back to USB, which we already have, this interface replaced USB. We’ll have output options for spdif/aes/bnc/i2s but the benefits then become limited to the internal part inside of the server. Which can still be considerable btw.
 
No then we’d be back to USB, which we already have, this interface replaced USB. We’ll have output options for spdif/aes/bnc/i2s but the benefits then become limited to the internal part inside of the server. Which can still be considerable btw.
Ok
Bnc out on dac card to bnc input on scaler..
 
The intended product will be modular and offer either Digital or Analog outputs. With the former, you can specify "standard" digital interface formats that still allow the use of any DAC and upsampler you like. You can also specify the new Taiko interface format which will limit the connectivity to manufacturers that have also implemented this new format. The analog-output DAC version, finally, is a closed system.

The Modular Daughter Board will provide one specific function from an array of options such as:
1. A DAC with Analog output
2. Generic S/PDIF, AES/EBU, I2s, or other digital formats
3. Custom output like MSB Pro ISL or Dual AES/EBU for DCS
4. New Taiko proprietary interface

As for the modular design, a few renders may help clarifying:

53251f07-796c-4538-b2b6-9b1aa5676c7f.jpeg

8b478bcb-e88f-4223-9c14-8fc683671098.jpeg

The “white block” is a solid machined piece of copper which contains and shields the optional output modules which in turn plug into the interface carrier board. In this render it contains the analogue output module.
 
im in Washington DC looking out on Pennsylvania Avenue and reading again about all things Taiko. Good to see this thread back on topic and read about what’s in the pipe line. Kudos Emile
 
Indeed Toslink should be able to perform competitively to SPDIF but its implementation is often done in the cheapest possible way.

A further consideration is sample rate support which per interface is limited to:

Toslink : 96KHz, no native DSD
SPDIF : 192KHz, no native DSD
AES/EBU : 192KHz, no native DSD
USB : currently supports but technically not limited to : 1536KHz, native DSD up to DSD2048

Then there is Ethernet ofcourse but that’s a different type of interface as it requires a “streamer / endpoint” to reside inside your DAC, to quote @MarcelNL , “I have yet to hear a good implementation”. There are more advanced interfaces running on Ethernet, like Ravenna, but these are specialised interfaces and exclusive to just a handful of DACs, these can also be used for high bandwidth multichannel systems.

The new kid on the block will be our own interface launching in the next few months, designed from scratch, both hardware, firmware and drivers, supporting any sample rate. Which is CPU direct over PCIe to I2S (“DAC language”). It can be clocked from both source as destination.

Why we went through the elaborate process of designing this is all the “off the shelf” solutions, which are the “engineering blocks” everybody uses, are aging and development on those have stopped. Further to this all of them have limitations in one way or another, they’re standard building blocks needing specific surrounding circuitry to operate which you have no choice in but to follow. Things like jitter reduction, reclocking, very expensive clocks etc are really just bandaids to reduce the negative impact of the existing issues in these standard circuits/protocols.
I apologize but i may have mixed up my terminology of the input i was using. based on the sample rates supported i believe i was using the optical input of the spdif module and not toslnink as i was playing 192khz pcm without issue. thank you very much for the clarification and information about the best input from the taiko.
 
The intended product will be modular and offer either Digital or Analog outputs. With the former, you can specify "standard" digital interface formats that still allow the use of any DAC and upsampler you like. You can also specify the new Taiko interface format which will limit the connectivity to manufacturers that have also implemented this new format. The analog-output DAC version, finally, is a closed system.

The Modular Daughter Board will provide one specific function from an array of options such as:
1. A DAC with Analog output
2. Generic S/PDIF, AES/EBU, I2s, or other digital formats
3. Custom output like MSB Pro ISL or Dual AES/EBU for DCS
4. New Taiko proprietary interface
Just to make sure I’m understanding this as I follow along (not trying to nit pick) should 1) read a DAC with analog input? If not, not sure I understand,
 
Just to make sure I’m understanding this as I follow along (not trying to nit pick) should 1) read a DAC with analog input? If not, not sure I understand,
A DAC inside the Extreme with a analog output to feed a preamp.
 

About us

  • What’s Best Forum is THE forum for high end audio, product reviews, advice and sharing experiences on the best of everything else. This is THE place where audiophiles and audio companies discuss vintage, contemporary and new audio products, music servers, music streamers, computer audio, digital-to-analog converters, turntables, phono stages, cartridges, reel-to-reel tape machines, speakers, headphones and tube and solid-state amplification. Founded in 2010 What’s Best Forum invites intelligent and courteous people of all interests and backgrounds to describe and discuss the best of everything. From beginners to life-long hobbyists to industry professionals, we enjoy learning about new things and meeting new people, and participating in spirited debates.

Quick Navigation

User Menu